M. Bonaèiæ, Anaphoric figures in the translation of poetry - SRAZ XLVII-XLVIII, 23-36 (2002-2003)
UDC 82-1:81’25
Original scientific paper
Accepted for publication on 26 September 2003
Anaphoric figures in the translation of poetry
Mirjana Bonaèiæ
Department of English
University of Split
Translated texts provide evidence of linguistic processing and meaning representation, since
domains of meaning and reference are constructed during translation. Rather than simply
assess the translation from the perspective of the original, also, from the perspective of the
translation, we should understand the original. In poetic discourse, figurative effects play a
central role in world creation in poems. They are often produced through anaphoric reference
by means of pronominal and nominal expressions. The article shows how complex anaphoric
figures are prompted by language use in the process of translating a poem, and how they can
be accounted for under the cognitive linguistic theory of conceptual blending, in which reference has a specific dimension of structure.
1. Introduction
A poetic text world is constructed from a text which is dissociated from any immediate
context of situation. With apparently nothing before, after or outside it, framed, as it were,
by silence, the text is striking to look at. Its patterns of language invite the reader to
become absorbed in an interactive process of constructing a mental representation of a
poetic world all of its own. The poetic text raises specific questions about the possibilities
of meaning construction, since there is no context of situation for it to refer to and
constrain its meaning. Nevertheless, the language of such a text supports the creation of
a context, without which it would be impossible to derive a poetic discourse or conjure up
the illusion of a world.
Context creation in a poem rests on how the reader interprets the enabling patterns of
its language, particularly the referential expressions the poetic text brings into play. These
expressions function in unexpected ways contributing to the allusive nature of poetic
23
02-M.Bonacic-gg..pmd
23
16.4.2004, 20:06
M. Bonaèiæ, Anaphoric figures in the translation of poetry - SRAZ XLVII-XLVIII, 23-36 (2002-2003)
discourse. For instance, what appears to be anaphoric reference often occurs at the
beginning of poems. An example is Wordsworth’s use of the pronoun she in the first
stanza of one of his untitled poems:
A slumber did my spirit seal,
I had no human fears:
She seemed a thing that could not feel
The touch of earthly years.
The use of anaphoric reference implies previous mention, but there are no antecedents
in the text itself. By evoking a sense of shared knowledge, the text achieves the immediacy
of reference to an already existing world. The illusion of a situation is created and the
reader is dramatically thrust into the poem’s world.
In poetic texts, referential forms often produce complex figurative effects, which lend
themselves to variable interpretation. This article examines how such aspects of textual
cross-reference can be accounted for and, specifically, how they manifest themselves in
translation. It is in the process of translating a poem that they become conspicuous in two
ways: at the level of formal disparity between languages and at the level of variability
between particular semantic interpretations.
2. Theoretical issues
In traditional text studies, accounts of how stretches of language come to be interpreted
as texts focus on the principles of formal connectivity. According to Halliday and Hasan
(1976), an important means of cohesion within a text is co-reference (or cross-reference)
by backward-looking (anaphoric) and forward-looking (cataphoric) reference. In particular,
pronouns are the paradigm examples of expressions used by speakers to refer to “given”
entities. Where their interpretation lies within the text, they form textual or endophoric
relations. In contextual or exophoric relations, which play no part in textual cohesion, the
interpretation of referential forms lies in the immediate situational context or in shared
knowledge. The first and second person pronouns are typically exophoric, referring to
participants in the discourse, while the third person pronouns are typically endophoric,
referring to noun phrases in the text.
In poetic discourse, however, such clear distinctions are blurred. In Wordsworth’s
stanza, the first person pronoun I creates the illusion of a speaking voice, which is the
poetic convention of “lyrical voice”, requiring no situational context for its interpretation.
On the other hand, with no noun phrase in the text to refer to, the third person pronoun
she refers to somebody in an allusive context, a poetic world conjured up by the artifice of
lyrical voice. The traditional theory of direct reference either to a thing entirely outside the
text or to a linguistic unit in the text does not account for such blended effects.
24
02-M.Bonacic-gg..pmd
24
16.4.2004, 20:06
M. Bonaèiæ, Anaphoric figures in the translation of poetry - SRAZ XLVII-XLVIII, 23-36 (2002-2003)
With the development of discourse analysis, emphasis has shifted to discourse
processing. Brown and Yule argue that the distinction between exophoric and endophoric
reference is hard to draw, because in both cases the processor has a mental representation:
‘In the one case he has a mental representation of what is in the world, in the other he has
a mental representation of a world created by discourse’ (Brown and Yule 1983: 201). This
can also account for such cases as the poetic use of a third person pronoun without an
antecedent in the text. The pronoun functions deictically and a situation is inferred. The
reader has a mental representation of an allusive entity in the world created by poetic
discourse or, more precisely, by reader-text interaction. There is no need to formulate a
special theory of poetic discourse as deviant from normal discourse processing. Moreover,
such uses of anaphora in poems usually pose no difficulties to translators. In Wordsworth’s
poem, she translates into ona ‘she’ with the same effect of evoking a sense of fascinating
immediacy:
Moj duh zatvori èudan san;
Strah ljudski minu mene:
Ko stvar bje ona, što neznan
Dodir joj ljeta, mijene.
(Translated by Ivan Goran Kovaèiæ)
More recently, complex anaphoric phenomena have attracted the attention of cognitive
linguists. There are, for instance, cases of anaphoric reference that require splitting the
self in two parts. Lakoff and Sweetser (1994: ix-x) give the following example:
If I were you, I’d hate me.
If I were you, I’d hate myself.
In these sentences, me and myself refer to two different people, but as they are both
first person, they both should refer to the speaker. An equivalent problem can be observed
in the Croatian translation of the sentences.
Da sam ti, mrzio bih me.
Da sam ti, mrzio bih se.
Croatian has the same reflexive pronoun se for all persons so that its interpretation
depends on the subject with which it is co-referential. So in the translated sentences, me
and se refer back to two different people and to the same first person subject. The self is
split into two parts, one identified with the person spoken to and one remaining the
speaker.
In poetry, anaphoric reference can motivate the construction of complex figurative
effects of the split self and blended meanings, as it is shown below in the analysis of two
poems and their translations. Lakoff and Sweetser (1994) say that the semantic theories
25
02-M.Bonacic-gg..pmd
25
16.4.2004, 20:06
M. Bonaèiæ, Anaphoric figures in the translation of poetry - SRAZ XLVII-XLVIII, 23-36 (2002-2003)
using formal logic cannot deal with such co-reference to split antecedents. They claim
that the explanation of the problem is easy in the cognitive theory of mental spaces as first
developed by Gilles Fauconnier in 1985 (Fauconnier 1994).
Cognitive linguistics has brought about a major shift of interest. As it equates meaning
with conceptualisation or cognitive processing (Langacker 1990), its focus of analysis is
on cognitive constructs and dynamics, and its methods extend to contextual aspects of
language use, which means studying “full discourse”, language in context and inferences
actually drawn by participants in an exchange (Fauconnier 1999). Even meanings assigned
to isolated sentences are obtained by building local, minimal contexts, which shows the
impossibility of a straightforward, context-independent pairing of linguistic form and
meaning configurations, assumed by standard formalisations. The basic claim is that “it is
only in rich contexts that we see the full force of creative on-line meaning construction”
(Fauconnier 1999: 97). In particular, when uses of language that are traditionally viewed as
deviant are treated as central and widespread, understanding the complex linguistic
organisation involved leads to the study of local conceptual configurations or
interconnected “mental spaces”. As defined by Fauconnier (1994: 1-2), interconnected
mental spaces are “domains that we set up as we talk or listen, and that we structure with
elements, roles, strategies, and relations”. Although they are not part of the language
itself, language does not come without them. In fact, language is “a powerful means (but
not the only one) of specifying or retrieving key aspects of this cognitive construction”
(Fauconnier 1994: xxxvii). Particular linguistic forms (“space builders”) prompt mental
spaces and others help to set up and identify elements in them. However, linguistic
expressions do not refer to space elements. Rather, reference goes from the constructed
elements in mental spaces to the objects referred to, which “implies metaphysical choices
about what can be referred to (objects, individuals in possible worlds, etc.)” (Fauconnier
1994: 168). For any stretch of thought, typically mediated by language, a multitude of
mental spaces are not only set up but interconnected. Reference, inference, and structure
projection operate on the basis of connections or cross-domain functions. A simple example
is when in talking about Dorian Gray we build two mental spaces, one for Dorian Gray and
the other for the portrait of Dorian Gray, and link them by only one function of identity,
although from an objective standpoint there is no identity at all, and even from a subjective
standpoint the model and his portrait can differ as much as we like (Fauconnier 1994).
Another widely applicable cognitive operation is blending. Conceptual blending
“matches two input spaces through a partial cross-space mapping and projects selectively
from both inputs into a third space, the blend, which gets elaborated dynamically”
(Fauconnier 1999: 102). The process of mapping exploits and further develops shared
schematic structure so that this common structure is elaborated in a fourth space, the
generic space. The four spaces constitute a “conceptual integration network” (Fauconnier
and Turner 1998). Fauconnier’s example is “trashcan basketball” (or dustbin basketball),
a game invented by students in a dormitory. The inputs would be the game of basketball
and the dormitory situation. The mapping would link elements in both spaces (the ball to
26
02-M.Bonacic-gg..pmd
26
16.4.2004, 20:06
M. Bonaèiæ, Anaphoric figures in the translation of poetry - SRAZ XLVII-XLVIII, 23-36 (2002-2003)
the crumpled paper, etc.), with the initial generic space consisting of an object being
thrown into a container by someone. In the blend, which is run in the real world, the new
game gets elaborated with respect to the constraints of the dorm situation. Along the
continuum of blending, another example is the analogical counterfactual In France,
Watergate would not have hurt Nixon, which sets up “a blended situation with some
features of France and others of the U.S. and the Watergate scandal” (Fauconnier 1999:
103). Further along the continuum are cases like Ezra Pound’s metaphor Fear, father of
cruelty, where “the two domains (emotions/qualities and people/kinship) have no literal
overlap at all, and the projected shared schema is correspondingly abstract (causality)”
(Fauconnier 1999: 115).
Although this is a necessarily simplified description, it is easy to see that this theory
should be able to account for the complex phenomena of anaphoric reference mentioned
above. It should also account for the complexity of figurative effects prompted by anaphora
in poetic discourse. In this article, I first explore the use of reflexive pronouns in one of
Emily Dickinson’s poems with respect to the translation problem of formal disparity
between English and Croatian. Then I analyse anaphoric reference created by nominals in
a poem by Robert Frost with respect to both untranslatability between languages and
variability between particular interpretations of the poem.
3. Figurative reflexives
The first line of one of Emily Dickinson’s typically untitled poems reads:
1 Me from Myself – to banish –
Similar to Lakoff and Sweetser’s sentences, the line splits the self in two parts. Then
in the first line of the second stanza, the -self anaphor pronoun appears again, but this
time as the first person subject:
But since Myself – assault Me –
Under any standard grammar of English, this use is anomalous. In her account of
Dickinson’s frequent deviant use of reflexive pronouns, Margaret Freeman (1997) finds
that in the light of Fauconnier’s mental space theory, Dickinson’s apparently ungrammatical
-self anaphors are perfectly grammatical. Her analysis of many such instances in numerous
poems leads her to propose a rule for Dickinson, namely that “whenever a subject referent
in one (originating) space projects a mental space (target) via a trigger or space-builder, its
pronoun counterpart in the target space will take the corresponding -self anaphor form”
(Freeman 1997: 11).
Revealing as such principles of poetic composition may be, translation is more
concerned with problems that may vary from poem to poem. The process of translation
27
02-M.Bonacic-gg..pmd
27
16.4.2004, 20:06
M. Bonaèiæ, Anaphoric figures in the translation of poetry - SRAZ XLVII-XLVIII, 23-36 (2002-2003)
captures the process of conceptualisation, and the process is local. This means that the
problem of particular deviant usage may be dealt with differently in the translation of
different poems, irrespective of the general principle underlying all such instances of
ungrammaticality. In what follows I give an account of the poem and its translation, which
is therefore different from the analysis of the poem given by Freeman and which also
reveals the interpretative function of translation. The account draws on the theory of
conceptual blending (Fauconnier 1999).
In Croatian, as stated above, there is only one reflexive pronoun for all persons. It has
all case forms except the nominative (sebe, gen., acc.; sebi, dat., loc.; sobom, instr.), and
short, unstressed forms se and si. None of its case forms can be used (as poetic licence)
in the position of subject/agent without making the utterance incomprehensible. Moreover,
it is not a compound word containing an element such as self that could function
independently as a free nominal form and therefore be conceptually reflected in the reflexive
pronoun, as is the case in Dickinson’s idiosyncratic use of -self pronouns in the role of
subject/agent. Nor does Croatian have a noun semantically equivalent to self (except the
compound jastvo, made up according to German Ichheit and used exclusively in
philosophy). Croatian also has a reflexive possessive pronoun svoj ‘one’s own’, varying
according to gender and case and co-referential with the subject/possessor. It can be
elegantly combined with the first person pronoun in the nominative ja ‘I’ to form the
phrase svoje ja, ‘one’s I’, which has a compositional (anaphoric and deictic) referential
value. Otherwise, personal pronouns can be omitted, since Croatian marks distinctions of
person unambiguously in the verb. These linguistic features can be used in the translation
of Dickinson’s poem. The English me becomes the reflexive sebe or se, a form used in both
the accusative (1, 5, 13) and genitive case (13). The English compound reflexive myself
becomes the deictic-anaphoric phrase svoje ja, used in the genitive and instrumental
case: svog ja (1), and svojim ja (5). With the personal pronoun ja, the “I” of the poem (the
self), remaining verbally unexpressed, the fantasy of the split self is played against the
background of the undivided self, the “real” agent of the fantasy. Complex cross-reference
is constructed in the conceptual integration network of translational discourse.
The poem:
Emily Dickinson
1
2
3
4
5
6
7
8
9
Me from Myself – to banish –
Had I Art –
Invincible my Fortress
Unto all Heart –
But since Myself – assault Me –
How have I peace
Except by subjugating
Consciousness?
And since We’re mutual
28
02-M.Bonacic-gg..pmd
28
16.4.2004, 20:06
M. Bonaèiæ, Anaphoric figures in the translation of poetry - SRAZ XLVII-XLVIII, 23-36 (2002-2003)
10 Monarch
11 How this be
12 Except by Abdication –
13 Me – of – Me – ?
A Croatian translation:
1 Sebe iz svog ja – prognati –
[Me(refl.) from my(refl.) I (to)banish]
2 Da znam umijeæe –
[If (I)knew art/craft]
3 Neprobojna mi
tvrðava
[Impenetrable me(Dat./my) fortress]
4 Za cijelo srce –
[For all
heart]
5 Al svojim ja – se napadam –
But (by)my(refl.) I
me(refl.) (I)assault]
6 Pa kako imam mir
[So how have(I) peace]
7 Ako ne savlaðivanjem
[If not (by)overpowering/vanquishing(Instr.)]
8 Svijesti?
[Consciousness]
9 I jer smo uzajamni
[And because (we)are mutual
10 Vladar
[Ruler/monarch]
11 Kako to bude
[How this is(perfective/be)
12 Ako ne abdikacijom –
[If
not (by)abdication(Instr.)]
13 Sebe – od – sebe – ?
[Me(refl.) of me(refl.)]
Translated by Mirjana Bonaèiæ, 2003
From the mental “real-world” space of the speaking (lyric) voice, the space-builder
Da znam ‘If I knew’ (2) triggers a hypothetical space in which the role of “I” (the self) takes
on two values. These are elaborated in the generic space that can be defined as abstract
binary structure of opposites. The two parts of the binary pair are specified by the linguistic
forms sebe ‘me’ and svoje ja ‘my I’ (1), which are both related anaphorically to the first
29
02-M.Bonacic-gg..pmd
29
16.4.2004, 20:06
M. Bonaèiæ, Anaphoric figures in the translation of poetry - SRAZ XLVII-XLVIII, 23-36 (2002-2003)
person agent in the projected, hypothetical mental space, as well as deictically to the
detached (elided) ja ‘I’ of the implied speaker’s “real-world” space. As the discourse
develops, metaphorical cross-space mappings link the hypothetical input space with
several other input spaces in order to structure a complex blend. One input space is
prompted by the words srce ‘heart’ (4) and svijest ‘consciousness’ (8). It exploits the same
initial generic space and gets contiguously (metonymically) connected with the abstract
schematic opposition between emotion and reason. These blended elements are connected
with the hypothetical space elements, so that one part of the split self, sebe or se, is
associated with heart/feelings, and the other, svoje ja, with consciousness/thoughts.
Another input (source) space contains the role of ruler and the relation between ruler and
subject, with a schematic structure of the strategies of reign and subjection, and eventually
abdication. These space elements are set up by the words: savlaðivanje ‘overpowering’
(7), vladar ‘ruler/monarch’ (10), and abdikacija ‘abdication’ (12). Still another metaphoric
source space consists of elements, strategies and relations belonging to a frame of war
and specified by the words encompassing the entire poem: prognati ‘banish’ (1),
neprobojna tvrðava ‘impenetrable fortress’ (3), napadati ‘assault’ (5), mir ‘peace’ (6),
savlaðivanje ‘overpowering’ (7), and also abdikacija ‘abdication’ (12), e.g. after defeat.
All the input spaces are matched through dynamic cross-space mappings which exploit
the shared schematic structure of conflict between the opposing elements. The mappings
connect all the input spaces and the generic space through the relevant cross-domain
functions, and project from them selectively into the blend, which gets elaborated
dynamically. This means that the conceptual integration network, set up by the language
of the poem, is not static. Shifting projective links connect the elements in the spaces by
complex cross-reference as the roles of “monarch” and “subject” alternate between the
two warring sides of the self, reason and emotion, and the drama of inner conflict develops
in the figurative world of poetic discourse.
This dynamic elaboration of the conceptual blending process can hardly be captured
by depicting a diagram of separate blends, such as in Freeman (1997). In the first stanza,
the target conceptualisation of a desire to suppress all emotion is constructed from the
blend of some of the elements of the metaphorical input (source) spaces: prognati ‘(to)
banish’ sebe ‘me’/srce ‘heart’ from the figurative tvrðava ‘fortress’ of svog ja ‘my I’. The
fortress belongs to the self as the agent of the hypothetical venture, so that cijelo srce ‘all
heart’/emotion might become the expelled subject/be annihilated.
In the second stanza, however, other elements come into play. The blend is modified
by the contrastive space builder Al ‘But’ (5). In the temporal gap between the two stanzas,
the desired activity seems to have been initiated, since reason is now waging war with
emotion. This is indicated by “my I” (svojim ja) being instrumental in assaulting “me” (se),
although the background agent remains the elided “I” (the self), who now desires mir
‘peace’ (5-6). To achieve peace, the ruler of the fortress, svijest ‘consciousness’/reason
(with which svoje ja ‘my I’ is now clearly associated), must be overpowered (7-8).
In the third stanza, the blend is transformed again by the space builder I jer ‘And
since’ (9) introducing a conclusive argument. In the temporal gap between the stanzas,
30
02-M.Bonacic-gg..pmd
30
16.4.2004, 20:06
M. Bonaèiæ, Anaphoric figures in the translation of poetry - SRAZ XLVII-XLVIII, 23-36 (2002-2003)
the hypothetical war ends. The first person subject becomes plural, marked in the verb
smo ‘(we) are’ (9), and the two warring sides get united into a mutual monarch. However,
this is an impossible situation. The space builder Kako to bude ‘How this be’ (11) projects
absurdity into the blend. The only solution is that the binary monarch abdicates. The war
is won and lost. As it is impossible to distinguish between the two ruling parts of the self,
the final integrity of the self is achieved by annihilating the distinction altogether and by
giving up the idea of sovereignty. The desire to suppress either emotion or reason is
abandoned. The powerful figurative means to signify this is the use of the same anaphoric
form sebe in the last, concluding line: Sebe – od sebe –. “I” am all there is, and the
figurative play of different anaphoric expressions ends.
4. Nominals in anaphoric figuration
Anaphoric figures constructed by relations that nominals may enter into within the
framework of a poem are even subtler than those just described. The fact that they can go
unnoticed by the translator of the poem as well as by the analyst of the translation is
evidenced by my next example of two translations of the same poem. In a different
translation, what is highlighted as a case of untranslatability may become a rich source of
anaphoric figuration and a central moment of meaning construction.
In her article “Translating a poem, from a linguistic perspective”, Tabakowska tries to
“corroborate” intuitive interpretations and assessments by “a strictly linguistic analysis,
which is carried out in the cognitivist vein” (Tabakowska 1997: 25). The example she uses
is Robert Frost’s poem Nothing Gold Can Stay and its Polish translation by one of the
most prominent contemporary Polish poets and translators, Stanis³aw Barañczak. However,
the meaning of the expression “translating a poem” cannot be fully understood unless
one examines translation as an on-line cognitive process. A fuller account of how my own
translation differs from Barañczak’s is given in Bonaèiæ (2003). Here, my aim is to reconsider
a specific aspect of untranslatability from the perspective of conceptual blending theory.
The poem:
Robert Frost
Nothing Gold Can Stay
1
2
3
4
5
6
Nature’s first green is gold,
Her hardest hue to hold.
Her early leaf’s a flower;
But only so an hour.
Then leaf subsides to leaf.
So Eden sank to grief,
31
02-M.Bonacic-gg..pmd
31
16.4.2004, 20:06
M. Bonaèiæ, Anaphoric figures in the translation of poetry - SRAZ XLVII-XLVIII, 23-36 (2002-2003)
7 So dawn goes down to day.
8 Nothing gold can stay.
A Polish translation:
Wszystko, co z³ote, krótko trwa
[Everything which gold
briefly
lasts]
1 Z³otem przyrody – pierwsza zieleñ
[Gold(N) of-nature
first
green(N)]
2 Po niej – ju¿ nic prócz sp³owieñ, zbieleñ.
[After it
already nothing except fadings
whitenings]
3 Rozkwitu szczyt – to pierwszy listek,
[Of-efflorescence peak
this first
leaf(Diminutive)]
4 Lecz przez godzinê ledwie; wszystek
[But for
hour
only
all]
5 Zwykleje w liϾ natychmiast potem.
[Commons into leaf(Acc.) immediately
afterwards]
6 Tak Eden zszarza³ nam w zgryzotê
[So Eden grayed
us(Dat.) into worry(Acc.)]
7 Tak œwit nam blaknie w œwiat³o dnia.
[So dawn us(Dat.) fades
into light(Acc.) of-day]
8 Wszystko, co z³ote, krótko trwa.
[Everything which gold
briefly lasts]
Translated by Stanis³aw Barañczak (1992)
A Croatian translation:
Ništa zlatno ne ostane
[Nothing gold
not stays]
1 Prva je zelen zlato,
[First is green(N) gold(N)]
(The first greenness/greenery is gold)
2 Prirodu krasi kratko.
[Nature(Acc.) (it)adorns briefly]
3 Njen rani list je cvijet;
[Her early leaf is flower]
4 Al zaèas veæ je blijed.
[But in-moment already is pale]
32
02-M.Bonacic-gg..pmd
32
16.4.2004, 20:06
M. Bonaèiæ, Anaphoric figures in the translation of poetry - SRAZ XLVII-XLVIII, 23-36 (2002-2003)
5 List struni list u pad.
[Leaf shakes-loose leaf into fall(Acc.)]
6 Raj potonu u jad,
[Paradise sank
into grief(Acc.)]
7 Zora u dan se utrne.
[Dawn into day itself extinguishes]
8 Ništa zlatno ne ostane.
[Nothing gold(Adj.) not stays]
Translated by Mirjana Bonaèiæ (2001)
In Tabakowska’s opinion, Frost’s poem conveys “the concept of permanence
composed of impermanences” through a “conspiracy” of imperfective habitual verbs
(subsides, goes down) and ungrounded countable nouns (leaf, day), so that this concept
“confronts the translator with a true case of untranslatability” (Tabakowska 1997: 38). She
uses Langacker’s concept of bounding. Boundary is a conceptual contour that delimits
the extension of an entity or process and is a function of how we construe the conceived
entity or relation, e.g. countable versus mass nouns, and perfective versus imperfective
verbs (Langacker 1990). The conceptualisation of “count-in-mass”, expressed by using a
countable noun without an article, is more complex. An entity, although bounded, is not
grounded in any particular pragmatic context. It loses its individuality and is conceptualised
as a mass of individual entities. As Polish does not lexicalise the notion of (in)definiteness,
there can be no equivalent to the “conspiracy” between nouns and verbs. So Tabakowska
praises the “noteworthy” compensation strategy used by the Polish translator: “the two
gerunds in line 2, sp³owieñ and zbieleñ, are derived from perfective verbs, and used in the
plural, thus becoming conceptually equivalent to Langacker’s ‘imperfective habitual’”
(Tabakowska 1997: 38).
To my mind, it is precisely those expressions which Tabakowska regards as contributing
to the image of “permanent recurrence” that actually become an obstacle to a different
cognitive process of inference generation and conceptual blending. The Polish version
contains a large number of different noun phrases, such as sp³owieñ ‘fadings’ and zbieleñ
‘whitenings’ (2), rozkwitu szczyt ‘peak of blossoming’ (3), diminutive listek preceded by
deictic to ‘this’ in the expression to pierwszy listek ‘this first little leaf’ (3), and liœæ ‘leaf’
(5), as well as pronouns nic ‘nothing’ (2) and wszystek ‘all’ (4). An overemphasis on the
grammatical category of “count-in-mass”, whereby it is claimed that an entity such as a
leaf loses its individuality, makes one ask the important question: “What happens to the
individual leaf?” The use of so many different and additional words prevents the reader
from inferring the crucial meaning of change occurring in the same individual thing. The
language of the Polish version prompts dispersed mapping schemes which can hardly
structure a distinct anaphoric figuration such as is set up by the language of the original.
This version makes one try to translate the poem again in a very different way.
33
02-M.Bonacic-gg..pmd
33
16.4.2004, 20:06
M. Bonaèiæ, Anaphoric figures in the translation of poetry - SRAZ XLVII-XLVIII, 23-36 (2002-2003)
Perhaps, the translator is not confronted with a “true” case of untranslatability. This
is a case of untranslatability from a formal linguistic perspective, since it is a linguistic fact
that Slavonic languages have no articles and do not lexicalise (in)definiteness. From the
perspective of cognitive linguistics, which studies “full discourse” (Fauconnier 1999), in
the Croatian version the concept of count-in-mass is evoked by the anaphoric relationship
of list ‘leaf’ in line 3 to zelen ‘greenness/greenery’, the topic of the opening lines, and by
the repetition of the same word list ‘leaf’, used without any deixis in line 5. The first two
lines already prompt a general schema of nature and spring, to which the concept of leaf
is then linked. It is modified by a shared abstract structure (transience), so we conceptualise
list ‘leaf’ as referring to a mass of leaves that make up nature’s greenery as well as being
an individual leaf which changes into a different leaf.
In the Croatian version, the same word list ‘leaf’ is repeated three times: Njen rani list
je cvijet ‘Her early leaf is flower’ (3), List struni list u pad ‘Leaf shakes-loose leaf into fall’
(5). The translation draws attention to the repetition of the word leaf in Frost’s poem: Her
early leaf’s a flower (3), Then leaf subsides to leaf (5). The repetition of the same word
evokes the concept of “difference in the sameness”, as in the well-known verse by Gertrude
Stein, A rose is a rose is a rose is a rose. The repetition makes one suspect that “each time
the same word returns, it means something slightly different, so the expression gives the
impression of saying something that is semantically rich and therefore ambiguous” (Eco
1976: 270). In the Polish version, the enjambment between lines 4 and 5 and the use of the
indefinite pronoun wszystek ‘all’ (4), instead of the same word for “leaf”, make such
instantaneous and rich meaning construction impossible. In Frost’s line (5) and in its
Croatian version likewise, the implied meaning of internal change in the thing itself
combined with downward movement is just as important as the concept of permanent
recurrence or even more so.
The semantic richness of the line is derived from anaphoric figuration. In the framework
of the entire poem, a conceptual network of complex mapping and blending is constructed,
of which only a glimpse can be caught in this short article. In line 5, the first occurrence of
the word leaf, or (Croatian) list, motivates the reader to match the input space triggered by
leaf or list anaphorically with the input spaces constructed before (lines 1 and 3): one
contributed by gold or zlato and the other by flower or cvijet. The cross-space mapping
is partial, since only some qualities from the frame of gold (a special hue of gold that can
be attributed to early leaves, preciousness) and from the frame of flower (beauty, value
attributed to flowers) are projected into the blend, which is elaborated by the schematic
structure of yet another metaphorical mapping from the source, spring, onto the target,
youth. The relation between the first and the second occurrence of the word leaf, or list,
is marked by a verb, the English subsides and the Croatian struni ‘shakes loose’, which
both contain in their semantic structure the concept of decrease or loss and the concept
of downward movement (in the Croatian version of the line, further verbalised as pad
‘fall’). This concept is a prototypical orientational metaphor for deterioration and
destruction. It is projected onto the blend in which the second occurrence of leaf is now
34
02-M.Bonacic-gg..pmd
34
16.4.2004, 20:06
M. Bonaèiæ, Anaphoric figures in the translation of poetry - SRAZ XLVII-XLVIII, 23-36 (2002-2003)
specified by the absence of all previous qualities and anaphorically matched with the
input spaces constructed before (lines 2 and 4). The identity of leaf is split in two parts.
The entire integration network associated with the discourse of this line is dynamic. It
forces not only a schema of the transition from purity to decay but also a generic schema
at a higher level by highlighting the causality in time between the two states. Line 5, which
concludes the first part of the poem, is not purely descriptive but emphatically evokes a
new meaning of inner change the cause of which is in the thing itself. Nor are lines 6 and
7 only similes in the mode of pastoral convention. A semantically rich discourse is created
by the relationships of cause and consequence, reason and conclusion. It is the imminence
of inner change that brings about the events described in the lines. Thus, the possibility
of multiple interpretations brings to life a poetic world which is much more complex than
the overall image of permanent recurrence.
5. Conclusion
In conclusion, I suggest that an account of language use in translation can reveal
how figurative imagination constitutes the way we understand poems. Rather than simply
assess the translation from the perspective of the original, also, from the perspective of
the translation, we should understand the original. In poetic discourse, complex figurative
effects are produced by anaphoric reference, and their function in world creation in poems
can be accounted for in the light of the cognitive linguistic theory of mental spaces and
conceptual blending, in which reference has a specific dimension of structure. As
conceptual integration theory has been concerned so far mostly with single sentences,
analysis of full discourse within the framework of a poem and its translation can highlight
the dynamic process of conceptual blending, and therefore contribute to a wider application
of this theory.
REFERENCES
BONAÈIÆ, Mirjana (2003). “Spoznaja i prevoðenje” (Cognition and translation). In Stolac,
D., Ivanetiæ, N. and Pritchard, B. (eds.) Psiholingvistika i kognitivna znanost u
hrvatskoj primijenjenoj lingvistici. Rijeka: HDPL (in print).
BROWN, Gillian, and YULE, George (1983). Discourse Analysis. Cambridge: Cambridge
University Press.
ECO, Umberto (1976). A Theory of Semiotics. Bloomington, Ind.: Indiana University Press.
FAUCONNIER, Gilles (1994). Mental Spaces. Cambridge: Cambridge University Press.
(First published in 1985. Cambridge, Mass.: MIT Press.).
FAUCONNIER, Gilles (1999). “Methods and generalizations”. In Janssen, T. and Redeker,
G. 1999, 95-127.
35
02-M.Bonacic-gg..pmd
35
16.4.2004, 20:06
M. Bonaèiæ, Anaphoric figures in the translation of poetry - SRAZ XLVII-XLVIII, 23-36 (2002-2003)
FAUCONNIER, Gilles, and TURNER, Mark (1998). “Conceptual Integration Networks”.
Cognitive Science 22(2), 133-187.
FREEMAN, Margaret (1997). “Grounded spaces: deictic -self anaphors in the poetry of
Emily Dickinson”. Language and Literature 6(1), 7-28.
HALLIDAY, M.A.K. and HASAN, Ruqaiya (1976). Cohesion in English. London:
Longman.
JANSSEN, Theo, and REDEKER, Gisela (eds.) (1999). Cognitive Linguistics: Foundations, Scope, and Methodology. Berlin: Mouton de Gruyter.
LAKOFF, George, and SWEETSER, Eve (1994). “Foreword”. In Fauconnier, G. 1994, ix-xvi.
LANGACKER, Ronald (1990). Concept, Image, and Symbol. Berlin: Mouton de Gruyter.
TABAKOWSKA, El¿bieta (1997). “Translating a poem, from a linguistic perspective”.
Target 9(1), 25-41.
ANAFORIÈKE FIGURE U PREVOÐENJU POEZIJE
Prevedeni tekstovi pružaju podatke o jeziènom razumijevanju i o predoèivanju znaèenja,
buduæi da se znaèenjska i referencijska podruèja grade tijekom prevoðenja. Prije nego jednostavno
vrednovati prijevod s gledišta izvornika, treba i iz perspektive prijevoda razumjeti izvornik. U
pjesnièkom diskursu, figurativni uèinci igraju središnju ulogu u stvaranju svijeta pjesme, a èesto
nastaju iz anaforièkih odnosa pomoæu pridjevnih i imeniènih izraza. Analiza pokazuje kako se
složene anaforièke figure stvaraju jeziènom uporabom u procesu prevoðenja pjesme i kako se mogu
obrazložiti u okviru kognitivne lingvistièke teorije pojmovnog stapanja, u kojoj je referencija posebna
dimenzija strukture.
36
02-M.Bonacic-gg..pmd
36
16.4.2004, 20:06